Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of All Answers.
In addition to the higher psychiatric morbidity among prisoners than the general population, it has also been identified that prisoners have higher rates of comorbidity. In Australia, general population studies have identified that in the past year the majority (80%) of individuals in the community did not have a psychiatric disorder (Teesson et al., 2009). Among those who had a psychiatric disorder a comorbidity disorder was rare as, 74.5% had one disorder, 22.0% had two disorders and 3.5% three disorders. In comparison, a large-scale Australian-based study conducted with male reception and sentenced prisoners found that the majority (61%) of prisoners had at least one psychiatric disorder (Butler et al., 2011). A comorbid disorder was also prominent among the sample with 25% of prisoners with a mental illness having a co-occurring substance use disorder. Among prisoners with a mental illness, the most prominent configuration of a comorbidity was for an affective and anxiety disorder (73%), followed by a schizophrenia spectrum disorder and an anxiety disorder (67%), or schizophrenia spectrum and affective disorder (51%). A recent systematic review and meta-regression study conducted by Fazel and Seewald (2012), also highlighted the high rates of comorbidity among prisoners. The authors identified that a substantial proportion of prisoners (20.4 to 43.5%) with a mental illness had a comorbid substance use disorder. A comorbid substance use disorder was more variable among prisoners with a schizophrenia spectrum disorder ranging from 13.6 to 95.0%, or an affective disorder ranging from 9.2 to 82.5%. The heterogeneity in research findings likely reflects differences in methodologies utilised as well as potentially real differences in prevalence rates between different types of prisoners (i.e., reception and sentenced) as well as between countries.
Despite considerable differences in how prevalence rates are operationalised, assessed and reported in the literature, the voluminous literature base has established that although prisoners suffer from the same spectrum of psychiatric disorders as the general population, the frequency, intensity and comorbidity of all types of psychiatric disorders are more prevalent among prisoners when compared to the general population. While it has been estimated that in the general population, one in two individuals will experience a schizophrenia spectrum, affective, anxiety, personality or substance use disorder, in the preceeding 12 months (Jablensky et al., 2000; Slade, Johnston, Oakley-Browne, Andrews, & Whiteford, 2009; Slade, Johnston, Teesson, et al., 2009), these rates were elevated (up to 80%) among prisoners (Butler et al., 2006; Fazel & Danesh, 2002; Short et al., 2010)
Regarding specific disorder clusters, there is substantial evidence that the prevalence of schizophrenia spectrum disorders, including schizophrenia, is at least three times higher among prisoners when compared to the general population. The literature has also consistenly demonstrated that all affective disorders, excluding dysthymic disorders among sentenced prisoners, are more prevalent amongst males incarcerated in prison than their community dwelling counterparts. Collectively the literature suggests that trauma and stress related disorders are more prominent among offenders than non-trauma and stress related disorders; finding that almost every second prisoner has experienced a form of anxiety disorder in the preceding 12 months. While personality and substance use disorders are relatively rare in the general population, personality and substance use disorders are the two most common diagnosis among male prisoners (Butler & Allnutt, 2003; Mullen et al., 2003). Taken together, the literature has established that a higher proportion of offenders entering the prison system as either remanded or sentenced prisoners (i.e., reception prisoners) have a psychiatric disorder rather than sentenced prisoners. Comorbidity is also more common among prisoners, as while one in six people (15.6%) in the general population have a primary and at least one co-occurring disorder, among prisoners at least one in four prisoners (24.6%) have a co-occurring disorder.
Even though the prevalence of psychiatric disorders is well recognised as being substantially higher among prisoners when compared to the general population, little is known about the long-term diagnostic stability of these psychiatric disorders among prisoners. Investigations of diagnostic stability have mainly excluded offenders and prisoners and have focused attention almost exclusively on the general population and community clinical settings. This section of the literature review discusses the importance of diagnostic stability, with attention drawn to the common methods employed in assessing diagnostic stability and reasons for diagnostic instability. This is followed by a review of the literature conducted with community populations that has investigated the diagnostic stability for schizophrenia spectrum, affective, anxiety and personality disorders. Lastly, common methodological limitations in community based diagnostic stability studies are explored.
Long-term stability of diagnoses is important in clinical practice as treatment options and prognosis for the patient largely hinge on the individual being correctly diagnosed (Chang et al., 2009; Whitty et al., 2005). For example, individuals diagnosed with schizophrenia would need different long-term pharmaceutical and psychiatric treatment than individuals diagnosed with drug induced psychoses where symptoms are likely to be absent after the acute phase of illness has passed (Whitty et al., 2005).
Despite the importance of valid and reliable diagnoses, these constructs are particularly difficult to evaluate, as there are limited or no biological markers to diagnose psychiatric disorders, precluding the ability of external validation. Furthermore, in clinical practice clinicians will observe patients’ symptoms longitudinally, seek collateral information from multiple sources, and consider differential diagnoses as part of their diagnostic formulation. This process is inevitably exposed to limitations regarding the information available to the clinician, as well as human error, which can reduce the accuracy of the diagnosis formulated. This circumstance exists albeit psychiatric diagnostic systems, including the ICD-10 (World Health Organisation, 1992) and DSM-5 (American Psychiatric Association, 2013) providing diagnostic criteria for psychiatric disorders. Consequently, validity and reliability of diagnoses is often evaluated by the stability of an individual’s diagnosis longitudinally (diagnostic stability), such as the consistency of the onset diagnosis with a follow-up diagnosis (Whitty et al., 2005). Diagnostic stability in this manner is a useful index for assessing the validity and reliability of diagnoses, as high diagnostic stability implies that the psychopathological or pathophysiological process is consistent across presentations (Fennig, Kovasznay, Rich, & et al., 1994).
Diagnostic stability can be calculated using three different methods including prospective consistency, retrospective consistency and 75% agreement across all evaluations (Schwartz, Fennig, Tanenberg-Karant, & al., 2000). Prospective consistency, otherwise referred to as positive predictive value (based on diagnosis being gold standard) is calculated
as the proportion of individuals that retain the same diagnosis in the last assessment as was given in the first assessment. Retrospective consistency is comparable to sensitivity and is calculated as the proportion of individuals that retain the same diagnosis in the first evaluation as was given in the last evaluation. Lastly, 75% agreement across all evaluations is calculated as the proportion of offenders who on at least 75% of assessments receive the same diagnosis.
Beyond the implications for the patient’s mental health, diagnostic stability using the methods described can have important implications from a public policy, training and research standpoint. This is because individuals who have a stable diagnosis are more likely to be true cases and place higher demands on mental health services. In contrast, diagnostic instability of disorders could lead to inappropriate service planning or resource allocation recommendations (Baca-Garcia et al., 2007). Therefore, it is also important to consider that diagnostic instability can occur for a number of reasons including observation, criterion, information or subject variance (Spitzer, Endicott, & Robins, 1975). Observation variance occurs when clinicians interpret the same stimuli in different ways or criterion variance where two clinicians use different criteria to diagnosis a disorder. Information variance occurs when the diagnosis shifts over time due to additional information becoming available or that previously gathered information is interrupted in a different manner at a follow-up assessment. Lastly, subject variance happens when there are actual changes in the patient’s symptomatology or whether clarity of symptoms occurred due to the patient’s response to treatment.
In studies investigating diagnostic stability, it is often not possible to ascertain on a case by case basis which type of error variance influenced diagnostic instability. Furthermore, in epidemiology diagnostic instability is typically attributed to procedural unreliability. However, this assumption is perhaps more pertinent for short-term rather than long-term follow-ups and dependent on the specific psychiatric disorder under investigation. In short-term follow-up studies, diagnostic instability should be low as changes in the patient’s symptomatology over short periods of time should be rare. In addition, short or long-term diagnostic instability for disorders such as schizophrenia spectrum and personality disorders should also be low, given these disorders are considered life-time disorders (American Psychiatric Association, 2013). Therefore, in these examples, concluding that diagnostic instability is due to observation, criterion or information variance would be reasonable. In contrast in long-term studies investigating the diagnostic stability of affective or anxiety disorders, it would be reasonable to assume that diagnostic instability is more likely to result from subject variance (i.e., a true change in symptomatology) rather than procedural unreliability. This premise reflects that individuals who are afforded psychiatric treatment for an affective or anxiety disorder may experience a decrease in symptoms or experience remission. A short-term follow-up evaluation that occurs before the patient experiences a reduction in symptoms, should have high prospective consistency. Conversely, when the follow-up evaluation occurs after successful treatment it would be reasonable to presume that the patient may have relapsed or was seeking psychiatric services due to the emergence of new symptoms pertaining to a different disorder. Hence, when considering diagnostic instability, it is important to take into consideration the length of the study and the psychiatric disorder under investigation.
Even though schizophrenia spectrum disorders are severe mental illnesses that are considered lifetime disorders (Fazel & Seewald, 2012; Fazel & Yu, 2011), studies have mainly focused on investigating the prospective diagnostic stability of schizophrenia. This is due in part, to the fact that the nature and severity of symptoms of psychosis can fluctuate over time. These studies have mainly followed up patients with first episode psychosis (Veen, Selten, Schols, & et al., 2004; Whitty et al., 2005) or a confirmed schizophrenia diagnosis at time one, that have required hospitalisation in a psychiatric hospital (Fennig et al., 1994; Kendler, Gruenberg, & Tsuang, 1985; Richard, Swann, & Burt, 1996; Tsuang, Woolson, Winodur, & et al., 1981). Among these studies, high prospective consistency for schizophrenia have been reported, ranging from 78.1% to 96% where six months to 40 years elapsed between the onset diagnosis and follow-up diagnosis and sample sizes ranged from 75 to 936. When study parameters are expanded to include multiple settings, such as combining inpatient, outpatient and emergency department, as well as including individuals diagnosed with non-psychotic disorders, diagnostic stability reduces to 68.6% (Baca-Garcia et al., 2007). This suggests that the ecological prospective diagnostic stability is only moderate.
Fewer studies have investigated retrospective consistency for schizophrenia. Among the studies conducted, retrospective consistency has been found to be lower than prospective consistency ranging from 45% to 73%, when follow-ups occurred between four to 12 years after the onset diagnosis (Baca-Garcia et al., 2007; Richard et al., 1996; Schwartz et al., 2000). The lower retrospective consistency when compared to prospective consistency for schizophrenia suggests that a sizeable number of individuals will take several years to be diagnosed with schizophrenia after contacting mental health services. There was also evidence that diagnostic stability often decreased as the length of the follow-up period and number of evaluations increased. This was illustrated by Schwartz et al. (2000) where the retrospective consistency for schizophrenia fell from 73% when the six and 24 months’ evaluations were compared, to 55% when the baseline and 24 month evaluations were compared. Consequently, it may take up to two years, for many individuals to be diagnosed with schizophrenia. This may account for the higher diagnostic stability initially which then reduces and stabilises over subsequent years. This premise also draws support from the lower level of retrospective consistency reported by Schwartz el al. (2000) for evaluations two years apart being comparable to the rates reported by Baca-Garcia (2007; 55% vs. 45.9% respectively) where the minimum timeframe between the onset and follow-up evaluation was approximately 12 years.
Less research attention has been afforded to investigating the diagnostic stability of specific schizophrenia spectrum disorders. However, prospective consistency has been found to be low for paranoid schizophrenia, 46.4% to 53.7% (Baca-Garcia et al., 2007; Chang et al., 2009; Kendler et al., 1985; Tsuang et al., 1981), hebephrenic schizophrenia, 40.9% (Kendler et al., 1985), residual schizophrenia 49.3% (Baca-Garcia et al., 2007), delusional disorder, 34.5 to 41.7% (Baca-Garcia et al., 2007; Whitty et al., 2005) and schizoaffective disorder, 50% (Laberge & Morin, 1995). Collectively, these findings suggest that in clinical practice symptoms pertaining to specific schizophrenia spectrum disorders vary considerable over the course of the illness (Chang et al., 2009; Whitty et al., 2005). Possibly due to the characteristics of the patient interacting together with longitudinal clinical changes resulting in diagnostic shift (Chang et al., 2009; Richard et al., 1996). Unclear delete?
Unlike schizophrenia spectrum disorders that are more often considered as lifetime disorders, affective and anxiety disorders are usually conceptualised as episodic (Donovan, Glue, Kolluri, & Emir, 2010; Vittengl et al., 2007). Overall, affective disorders have been identified to have moderate stability (54.9 to 78.3%; Baca-Garcia et al., 2007; Tsuang et al., 1981). However, specific affective disorders generally have lower prospective consistency with major depressive disorder ranging from 40.3% to 75% and bipolar ranging from 35.4% to 56% (Andreasen et al., 1981; Baca-Garcia et al., 2007; Bromet, Dunn, Connell, Dew, & Schulberg, 1986; Fendrich, M., Warner, & L., 1990; Rice et al., 1986; Rice, Rochberg, Endicott, Lavori, & Miller, 1992; Tsuang et al., 1981). Prospective consistency for anxiety disorders has been reported to be even lower than the rates reported for affective disorders. Most studies have identified low prospective consistency for phobic disorder (33 to 52%), obsessive-compulsive disorder (19.2% to 66%), panic disorder (35 to 66%) and generalised anxiety disorder (15 to 29%; Andreasen et al., 1981; Baca-Garcia et al., 2007; Fendrich et al., 1990; Nelson & Rice, 1997; Rice et al., 1992). A general trend in diagnostic stability was identified for affective and anxiety disorders. Studies conducted in one clinical setting with a short follow up period of up to two years generally reported moderate diagnostic stability and as the number of sites and length of follow-up increased, diagnostic stability decreased (ref).
Similar to schizophrenia spectrum disorders, personality disorders have generally been characterised as lifetime disorders, as traits and behaviours developed during childhood and adolescence continue throughout adulthood and are relatively resistant to change (American Psychiatric Association, 2013). Early studies investigating the diagnostic stability of personality disorders supported this assumption as personality disorders were identified as being mainly stable over time (Carpenter, Gunderson, & Strauss, 1977; Grinker, Werble, & Dryre, 1968; Gunderson, Carpenter, & Strauss, 1975; Maddocks, 1970; Robins, Gentry, Munoz, & Marten, 1977; Skodol, Buckley, & Charles, 1980; Werble, 1970). Personality disorders were also associated with ongoing poor functioning across multiple domains and symptomatic impairment (Grilo, McGlashan, & Oldham, 1998). However, these studies were restricted to investigating the stability of borderline (Carpenter et al., 1977; Grinker et al., 1968; Gunderson et al., 1975; Skodol et al., 1980; Werble, 1970) or dissocial (Maddocks, 1970; Robins et al., 1977) personality disorders. The studies also had small sample sizes (n = 24 to 59), were conducted in either a single inpatient or outpatient clinic and most had relatively short follow up periods (i.e. less than three years). Combined these methodological limitations likely inflated diagnostic stability.
Numerous studies investigating the diagnostic stability of borderline personality disorder post the introduction of the DSM-III (American Psychiatric Association, 1980) addressed many of the methodological limitations of early studies. This included increasing the follow-up period (i.e. up to 20 years) and larger sample sizes, although many were still often restricted to one clinical setting. Collectively the studies provided evidence that borderline personality disorder was less stable over time as many individuals, experienced fewer symptoms and had improvements in social and occupational outcomes (Bardenstein & McGlashan, 1988; McGlashan, 1984; Paris, Brown, & Nowlis, 1987; Paris, Nowlis, & Brown, 1988; M. H. Stone, 1987; M. H. Stone, Hurt, & Stone, 1987). Low diagnostic stability has also been identified for dissocial personality disorder (42.9 to 58.8%; Black, Baumgard, & Bell, 1995; Helzer, Spitznagel, & McEvoy, 1987; Perry, 1988; Perry, Lavori, Cooper, Hoke, & O’Connell, 1987; Vandiver & Sher, 1991) and all other personality disorders (43 to 56%; Bernstein et al., 1993; Klein & Ferro, 1997; Loranger, Sartorius, Andreoli, & et al., 1994; McDavid & Pilkonis, 1996; Orlandini, Fontana, Clerici, & et al., 1997).
Today there is substantial empirical evidence that has refuted the long-term stability of personality disorders, as more often personality disorders demonstrated low diagnostic stability ranging from 27.8% to 34.7% (Baca-Garcia et al., 2007; Grilo & McGlashan, 1999; Grilo, Sanislow, Gunderson, & et al., 2004; Lenzenweger, Johnson, & Willett, 2004; McDavid & Pilkonis, 1996). There is also emergent support for personality disorders being less enduring (Durbin & Klein, 2006; Shea et al., 2002), hybrids of trait-like attitudes and symptomatic behaviours (McGlashan et al., 2005), as well as being state-based (Reich, 2002). Fluctuation of personality disorder symptoms overtime, perhaps are due to maladaptive coping skills, since symptoms can manifest and abate in conjunction with symptoms of another psychiatric disorder.
Divergent methodologies have been used in the literature assessing diagnostic stability and it is likely that many community based studies have biased the results towards higher levels of stability then would be seen in ecological settings. First, sampling bias including recruiting participants from one mental health setting, such as a single outpatient setting, and drop-out were common problems with longitudinal studies. Second, evaluating clinical decisions made by the same clinician, clinicians not being blind to the purpose of the study and evaluating a single diagnostic cluster (such as psychotic disorders; Mojtabai, Susser, & Bromet, 2003; Rufino, Uchida, Vilela, & al., 2005; Schimmelmann, Conus, Edwards, & et al, 2005; Schwartz et al., 2000; Veen et al., 2004), while able to control for observer differences, overestimates diagnostic stability. Third, diagnostic procedures, including using semi-structured interviews or other diagnostic assessments that are not routinely used in clinical practice may not reflect the diagnostic stability of naturally occurring clinical decisions in ecological settings. Fourth, most studies had short follow-up periods (usually less than 3 years; Barkow, Heun, Wittchen, & et al., 2004; Grilo & McGlashan, 1999; Grilo et al., 1998; Grilo et al., 2004; Rufino et al., 2005; Schimmelmann et al., 2005; Schwartz et al., 2000; Veen et al., 2004) and were limited in the number of follow-up evaluations (usually 2 or 3 evaluations) (Grilo et al., 2004; Schimmelmann et al., 2005; Schwartz et al., 2000). The premise that the methodologies used in most community studies likely inflated diagnostic stability rates is supported by Baca-Garcia et al., (2007) that revealed that longitudinal ecological diagnostic stability was poor. Specifically, diagnostic stability was as low as 35% for disorders of adult personality and behaviour, 55% for affective disorders, to 69% for schizophrenia spectrum disorders. In this study the authors investigated the ecological diagnostic stability in a variety of community clinical settings including emergency departments, inpatient and outpatient settings.
Studies evaluating diagnostic stability among individuals with an offending history are required, as findings from community based studies may not be generalisable given the well-established differences in the profile of mental illness between offenders and the general community. Prisoners will more often have complex presentations as psychiatric morbidity, dual diagnosis and co-occurring substance abuse problems are more common among prisoners than the general population (Butler et al., 2011; Fazel & Seewald, 2012). Furthermore, compared with the general population, individuals with an offending history often have different pathways to care, are less likely to utilise psychiatric services, have a heightened risk of misdiagnosis, treatment noncompliance and are stigmatised which can effect perceptions of illness aetiology (Kinner, 2006; Williams, Skogstad, & Deane, 2001). Collectively these differences may affect the individual’s course and outcome, as well as the validity of diagnoses ascribed by clinicians.
Diagnostic stability of psychiatric disorders has been investigated substantially among the general population, however, there is a stark absence of studies conducted with prisoners. Among studies conducted there is considerable differences in the level of diagnostic stability reported for schizophrenia spectrum, affective, anxiety and personality disorders. The general trend identified is that diagnostic stability is high for all disorders when participants are recruited from one mental health setting, assessments are conducted by the same clinician and few follow-up evaluations over a period of less than two-years are conducted. However, as the number of mental health setting, clinicians, number of assessments, and length of follow-up period increases, diagnostic stability starts to drop. This pattern suggests that the methodology employed in many studies inflate the level of diagnostic stability beyond what would be seen in clinical practice. In ecological settings, it has been identified that diagnostic stability for schizophrenia is moderate and for affective and personality disorder diagnostic stability is low.
Studies evaluating diagnostic stability among individuals with an offending history are required, as findings from community based studies may not be generalisable, as there are well established differences in the profile of mental illness between offenders and the general community. Th differences in the profile of mental illness may result in clinicians encountering more difficulties in correctly diagnosing mentally ill offenders when compared to mentally ill community dwelling individuals. This gap in the literature should also be addressed because diagnostic stability has implications for clinical practice, public policy, training and research.
Even though psychiatric morbidity and co-occurring disorders are higher among prisoners when compared to the general population, there is no singular explanation for the higher prevalence rates. While, it was not a direct aim of this thesis to investigate the reasons for the higher prevalence rates, it is important to acknowledge that many competing and complex explanations have been provided. Some of these explanations relate to the evolving mental health system such as: deinstitutionalisation, a lack of adequate general psychiatric and specialist services or diversionary options in the community and more formal and rigid criteria for civil commitment (Gunn, Maden, & Swinton, 1991; Lamb & Weinberger, 1998; Victorian Auditor-General’s Office, 2014). Other factors pertain to experiences in the community such as homelessness, the reluctance of general psychiatric services to accept mentally ill patients from the courts and society’s intolerance of deviant behaviour by people with a mental illness (Coid, 2003; Kinner, 2006). In contrast, others have argued that the higher prevalence rates reflect heightened awareness amongst both professionals and the public (Ogloff, 2002). These factors have different historical underpinnings and an exploration of the chronological history is warranted as this will help to understand how the public mental health system has evolved over the last several decades. The review will also enable an understanding of the current mental health system that offenders will be obtaining treatment.
Prior to deinstitutionalisation, psychiatric care was almost entirely provided in long-stay stand-alone psychiatric institutions (i.e., asylums). During the mid-1960s, when the number of psychiatric beds in stand-alone institutions in Australia peaked at approximately 30,000 (Department of Health and Ageing, 2013), concerns were mounting in Australia and internationally about the ill treatment of people with a mental illness. The main themes included isolation, inhumane treatment and the practice of long-term detainment of mentally ill people, as more often individuals who became inpatients were never discharged (Ashley, 1922; Pollock, 1938). This practice was even more prominent among patients with a prior offending history or who were deemed a danger to society (Pollock, 1938). This occurred even though studies conducted pre-deinstitutionalisation shared the assertion that individuals with a mental illness were not more dangerous than other people (Ashley, 1922; Cloninger & Guze, 1970; Cohen & Freeman, 1945; Guze, Goodwin, & Crane, 1969; Guze, Woodruff, & Clayton, 1974; Pollock, 1938).
Between the 1960s and 1990s a systematic deinstitutionalisation movement occurred in Australia. The stand-alone psychiatric hospitals were progressively closed, reducing the number of psychiatric beds dramatically to 5,802 (33 per 100,000) by 1992-93 (Department of Health and Ageing, 2013). The closures of stand-alone psychiatric hospitals had a radical impact on Australia’s mental health system, as there were insufficient community resources to offset the reduction in psychiatric services previously provided by psychiatric hospitals (Australian Health Ministers, 2003). Inadequacies existed in outpatient services, community residential services and 24-hour psychiatric care facilities for severely mentally ill people requiring long term care (Australian Health Ministers, 1998). Without adequate service options, the system was no longer able to keep up with the demands of mentally ill individuals and services were grappling to meet even the highest priority needs of mentally ill individuals (Australian Health Ministers, 1992). Similar deficiencies were also apparent internationally, and it was during this time that deinstitutionalisation was implicated as a key factor in increasing the risks of mentally-ill individuals becoming incarcerated (Davis, 1992; Hodgins, 1992; Hodgins, Mednick, Brennan, & et al., 1996; Palermo, Smith, & Liska, 1991; Sosowsky, 1980; Torrey, 1997).
One of the most critical impacts of deinstitutionalisation was that large-scale closures of stand-alone psychiatric hospitals resulted in deficiencies in accommodation and treatment options for mentally ill people living in the community (Australian Health Ministers, 1992; Jemelka, Trupin, & Chiles, 1989; Laberge & Morin, 1995). Without affordable community housing options, mentally ill people were at risk of experiencing homelessness, which heightens the risk of incarceration (Belcher, 1988; Galea & Vlahov, 2002; Kushel, Hahn, Evans, & et al, 2005; Metraux & Culhane, 2006). Becoming itinerant also heightens the risk of mental status decompensation, due to disengaging with voluntary mental health services and ceasing to take prescribed medications (Belcher, 1988). For individuals diagnosed with schizophrenia, these risk factors place the individual at jeopardy of being in a psychotic state within the community and, coming into contact with police (Belcher, 1988).
Changes in civil commitment laws also came into effect during deinstitutionalisation resulting in more formal and rigid criteria for civil commitment, reducing the likelihood of mentally ill people receiving mental health services that they needed (A. A. Stone, 1978) dated reference. Additionally, the changes afforded people the opportunity to refuse treatment, resulting in more mentally ill individuals going untreated in the community (Lamb & Weinberger, 1998; Laub, Nagin, & Sampson, 1998; Teplin, 1994). This, together with living in the community where mentally-ill individuals could access substances that were unavailable in psychiatric hospitals, also increased the risks of offending. Untreated mentally ill people, especially those who also abuse substances, are at increased risk of violence, which increases the probability of the person becoming involved in the criminal justice system (Fulwiler, Grossman, Forbes, & et al., 1997; Hodgins et al., 1996; Kushel et al., 2005; Mulvey, 1994; Steadman, 1997; M. H. Stone, 1997; Swanson, Estroff, Swartz, & et al., 1997).
Mentally ill people with an offending history or those recently released from prison, were also less likely to access adequate general psychiatric, specialist services or diversionary options in the community, as there was a reluctance of general psychiatric services to accept this dually-stigmatised sub-group of mentally ill patients (Coid, 2003; Gunn et al., 1991; Jemelka et al., 1989; Kinner, 2006; Laberge & Morin, 1995; Lamb & Weinberger, 1998; Watson, Corrigan, & Ottati, 2004).. These factors, in conjunction with changes in police decision making practices and society’s intolerance of deviant behaviour by mentally ill people, increased the risk of mentally ill individuals becoming involved in the correctional system, especially for minor offences (Coid, 2003; Gunn et al., 1991; Lamb & Weinberger, 1998).
Taken together, these key factors have been used to argue that seriously mentally-ill individuals were being diverted from stand-alone psychiatric hospitals into the correctional system (Teplin, 1983). As a result the term “criminalisation of the mentally ill” was coined (Abramson, 1972) and prisons have been referred to as the new psychiatric hospitals that provide public psychiatric care to mentally-ill people (Torrey, 1995). Deinstitutionalisation has also been cited as causing the revolving door principle of repeat incarcerations among mentally-ill individuals (Hoge, 2007). Despite it being argued that deinstitutionalisation has increased the proportion of mentally ill people being incarcerated (Davis, 1992; Palermo et al., 1991; Torrey, 1995), empirical evidence does not always support this contention (Wallace et al., 2004). Explain Wallace as these findings are important crossing the period of deinstitutionalisation – a bit more detail. There has also been a lack of longitudinal studies that have investigated changes in offending patterns that have incorporated the period of deinstitutionalisation or included a control sample. One study addressed this problem and demonstrated there was a higher rate of criminal convictions among individuals diagnosed with schizophrenia post deinstitutionalisation, especially for violent and drug related offences (Wallace et al., 2004). Nonetheless, this rate was matched by a proportionately similar increase in offending in the general population. Therefore, the higher proportion of mentally ill people entering the correctional system was argued by Wallace and colleagues (2004) to be unlikely the sole result of deinstitutionalisation.
Irrespective of whether deinstitutionalisation is responsible for the higher rates of mental illness among prisoners it is clear that prior to deinstitutionalisation there was limited opportunity for seriously mentally ill individuals that were lifetime residents in stand-alone psychiatric hospitals to commit offences, be arrested or incarcerated. Therefore, living in the community post deinstitutionalisation exposed seriously mentally ill individuals not only with an opportunity to offend but also exposed the individual to a range of risk factors associated with offending.
Pressure mounted for governments to work together and commit to reforming the public mental health system, after it was recognised that there were substantial inadequacies within the public mental health system. This culminated with the introduction of the National Mental Health Policy (Australian Health Ministers, 1992)which was endorsed by all Health Ministers in 1992. This was the first time since Federation that the eight state and territory governments worked together to redress service development issues requiring nationally focused policies. As prior to the National Mental Health Policy, Australian public mental health services were fragmented, as each of the eight state and territory governments were exclusively responsible for the running of their respective jurisdictions’ mental health services.
Reforms centred on integrating mental health services into mainstream health care, to rectify some of the inadequacies of the mental health system by investing in expanding inpatient and community services (Australian Health Ministers, 1992). To achieve this end, a cohesive mental health program was developed to replace the services traditionally provided in stand-alone psychiatric hospitals. This included transferring the provision of acute psychiatric inpatient care into community based general hospitals and expanding community based care alternatives. During the first five years, Victoria led the reforms by undertaking extensive structural changes to the public mental health system, whereas other jurisdictions were slower to reform services (Australian Health Ministers, 1998). Similar reforms were also carried out internationally, however expansions of community based psychiatric beds and psychiatric services has not been able to keep up with the reduction of stand-alone psychiatric hospital beds or services (Fakhoury & Priebe, 2007; Saxena, Thornicroft, Knapp, & Whiteford, 2007). Since the reforms did not redress the accommodation and treatment problems, this has likely further exacerbated the problems encountered by mentally ill individuals, especially those with a serious mental illness and / or offending history.
At the same time the Australian mental health system was being reformed, several reforms were also being undertaken within the Victorian criminal justice and forensic mental health systems. Two changes that significantly improved the services and care of mentally ill people in the criminal justice system was the establishment of the Victorian Institute of Forensic Mental Health (Forensicare) in 1997 and the subsequent opening of the Thomas Embling Hospital (TEH) in 2000. As a statutory agency, Forensicare is responsible for providing adult forensic mental health services, research, training and professional education in the health and justice sectors in Victoria (Victorian Institute of Forensic Mental Health, 2016). The TEH is a 116-bed secure forensic mental health hospital that provides advanced clinical treatment and programs. While most patients passing through the hospital are transferred from the criminal justice system for psychiatric assessment and / or treatment, the largest proportion of patients detained in the TEH are forensic patients (Victorian Institute of Forensic Mental Health, 2016). In addition, other prison based mental health services, such as St Paul’s psychosocial unit at Port Phillip Prison have also been established. These services aim to provide multi-disciplinary care, treatment and rehabilitation for male prisoners requiring assistance and integration into the mainstream prison population or wider community on release (G4S Correctional Service, 2017).
Additionally, in July 2007 the Victorian Department of Justice established the Justice Health business unit. Justice Health is responsible for the planning and coordination of health services across police, courts, and Corrections Victoria to ensure an integrated and coordinated approach for health services (Corrections Prisons & Parole, 2017a). The unit was established to consolidate the health functions previously provided in collaboration between Corrections Victoria and the Prisoner Healthcare Unit, Department of Human Services. The establishment of Justice Health has helped ensure the provision of quality driven and streamlined services with a centralised governance model.
A second priority post deinstitutionalisation was to investigate the prevalence and impact of mental illness among Australians. As a result, three cross-sectional surveys collectively known as the National Survey of Mental Health and Wellbeing (i.e., Department of Health and Ageing, 2013; Department of Human Services, 2007) were conducted. The first two surveys were conducted within the adult general poulation in 1997, with the first survey investigating the prevalence and impact of high-prevalence disorders including depression, anxiety and substance use disorders (Australian Bureau of Statistics, 1998). The second survey investigated low-prevalence disorders such as psychotic disorders (Jablensky et al., 2000). As the first two surveys focused solely on adults, a third survey commissioned in 1998 focused on mental illnesses among children and adolescents (Sawyer et al., 2001).
Collectively, the national surveys provided an understanding of trends in mental health, as well as contemporary estimates regarding the prevalence of mental illness among Australians. Nonetheless, the initial surveys failed to recognise or include prisoners as a sub-group that often experiences higher rates of mental illness than individuals in the general population. Consequently, the need and demand for mental health services among prisoners remained unknown and estimates of the prevalence of mental illness among prisoners was obtained from smaller prison based studies. These studies often had small samples (Brinded et al., 2001; Gibson et al., 1999; Herrman et al., 1991) and were conducted at single sites (i.e., one prison; Ghubash & El-Rufaie, 1997; Guy, Platt, Zwerling, & Bullock, 1985; Krefft & Brittain, 1963; Smith, O’Neill, Tobin, Walshie, & Dooley, 1996; Teplin, 1994). The studies were also often limited to investigating the prevelance of a single disorder (i.e., schizophrenia disorder; Bøjholm & Strömgren, 1989), or within sub-groups of offenders, such as remanded (Andersen, Sestoft, Lillebæk, Gabrieisen, & Kramp, 1996; Birmingham, Mason, & Grubin, 1996; Brinded et al., 2001; Brooke, Taylor, Gunn, & Maden, 1996; Davidson, Humphreys, Johnstone, & Owens, 1995) or sentenced (Bland, Newman, Dyck, & Orn, 1990; Chiles, Cleve, Jemelka, & Trupin, 1990; Gunn et al., 1991) prisoners.
Despite methodological differences, collectively these studies brought heightened awareness to prisoners, by identifying that prevalence rates of schizophrenia spectrum, affective, anxiety, personality and substance use disorders, among prisoners surpassed that found in the general population. Nonetheless, the studies also likely underestimated the prevalence of mental illness among prisoners, as most studies would have excluded the most acutely mentally ill prisoners. Acutely unwell prisoners more often are unable to provide informed consent or are inelligible to participate because of being admitted to prison psychiatric units.
In 2007 and 2010 respectively, the two adult national surverys were replicated (Morgan et al., 2011; Slade, Johnston, Oakley-Browne, et al., 2009; Slade, Johnston, Teesson, et al., 2009) and similar to the first surveys, prisoners were entirely excluded. This occurred even though empirical evidence continued to draw attention to prisoners having disproportionately higher rates of psychiatric morbidity than the general popultion (Fazel & Danesh, 2002).
The two adult national surveys, however, did incorporate questions that asked participants whether they had been previously charged with an offence or incarcerated. This enabled the first study to identify that people with a history of incarceration were twice as likely to have a high-prevalence mental disorder in the previous 12 months (41.1% vs. 20.0%; Slade, Johnston, Teesson, et al., 2009). This included higher prevalence rates of affective (19.3% vs. 6.2%), anxiety (27.5% vs. 14.4%) and substance use (22.8% vs. 5.1%) disorders, when compared to those whom had never been incarcerated. Despite identifying that offenders had higher prevalence rates, no further comparisons were made between offenders and the general population including exploring whether mental health needs or service utilisation differed between the two groups. In the 2010 study investigating schizophrenia spectrum disorders, an even more stark absence of comparisons between the two groups was noted. The only commentary that related to offenders, was that 10.6% of participants with a schizophrenia spectrum disorder had been charged with an offence and 3.2% had been incarcerated during the year prior to the study (Morgan et al., 2011).
The lack of attention afforded to prisoners and offenders in the national studies has persisted even though there is a clear need for these sub-groups to be incorporated. Especially given that a sizeable proportion of mentally-ill offenders will continually transition between the community and prison, taking their mental health service needs with them. As each year most prisoners will serve relatively short sentences of less than 12 months and approximately half will be re-incarcerated within one year (Broadhurst, Maller, Maller, & Duffecy, 1988). These factors highlight that a substantial proportion of prisoners will require access to both community and prison based mental health services. Furthermore, that continuity of care between the community and prisons is important, especially during heightened periods of stress, such as on entering prison, while imprisoned and upon discharge from prison. By prisoners accessing appropriate mental health services the risk of repeated incarcerations may be reduced (Baillargeon et al., 2009; Fazel & Yu, 2011). Accessing appropriate mental health services in prison can also reduce the risk of self-harm and suicide while incarcerated (Fazel, Cartwright, Norman-Nott, & Hawton, 2008; Lohner & Konrad, 2007). Upon release, continuity of care might also reduce drug-related deaths and suicide (Bird, 2008; Kariminia et al., 2007; Pratt, Appleby, Piper, Webb, & Shaw, 2010). Despite the potential benefits of appropriate treatment, offenders have been identified as encountering difficulties in accessing suitable community mental health services, possibly due to the dual stigma of having a mental illness and a criminal history (Kinner, 2006). Furthermore, that high rates of co-occurring disorders often preclude offenders from accessing community-based treatment (Ogloff, 2002; Ogloff, Lemphers, & Dwyer, 2004)Ogloff, Talerski, Lemphers, Simmons & Wood, 2015. For those that obtain treatment, co-morbidity may complicate treatment options, as therapy for different disorders may be incompatible with one another (Adams & Ferrandino, 2008).
Despite this evidence, national surveys have continued to overlook this particularly vulnerable sub-group of the population with unique and often complex needs and have missed an opportunity to address this critical shortfall. Furthermore, no known study exists that directly compares lifetime mental health service utilisation between prisoners and those in the community who do not offend. As a result it remains unknown whether the higher rates of mental disorders among prisoners than the general population, translates into prisoners having a greater use of mental health services. Given the lack of large scale research attention afforded to prisoners, it is also probable that mental illnesses among offenders and prisoners has always been higher than among the general population. However, it has only been in the last few decades with heightened awareness from the increasing amount of research attention afforded to offenders, that it has become more widely recognised that mental illness prevalence rates among offenders far surpasses that seen in the general population.
Since the reforms to the public mental health system commenced, new priorities have been developed and incorporated based on emergent knowledge and evolving community expectations. One of the key objectives incorporated since 1998 has been to decrease the prevalence and severity of mental illness among Australians. To work towards this goal, the National Mental Health Strategy embraced an explicit population health approach that acknowledged there are a number of important determinants of mental health wellbeing (Australian Health Ministers, 1998, 2003). These largely encompass psychosocial and environmental factors such as education, employment, income, and access to community resources (Visher, La Vigne, & Travis, 2004). However, these government initiatives have largely been directed towards children, adolescents or adults in the community that have never been incarcerated. This has occurred despite offenders being a socially disadvantaged sub-group of the population that experiences these psychosocial and environmental risk factors at increased rates than the general population. Numerous studies have identified that offender’s grapple to obtain education, employment and stable accommodation (Greenberg & Rosenheck, 2008; Martell, Rosner, & Harmon, 1995; Michaels, Zoloth, Alcabes, & et al., 1992; Solomon & Draine, 1995). As a result, offenders have minimal community and social supports (Laberge & Morin, 1995), limited economic security, and a lack of structure in their lives which increases their risks of being incarcerated for minor offences (Lamb & Grant, 1982). In addition, mentally ill offenders have been identified as being more likely to be victimised INCLUDE SHORTS ARTICLE or to victimise strangers (Martell et al., 1995).
Importance has also been placed on destigmatising mental illness and developing services that incorporate the complete gamut of mental health services from mental health promotion, prevention through to treatment (Australian Health Ministers, 2003). In the first instance, mental health promotion and prevention efforts aim to stop mental illnesses from developing, whereas effective treatment aims to decrease the duration of mental health symptoms. Multi-modal treatment options have become a central focus to reduce the severity of mental illness and to address disablement that mentally ill individuals experience in personal, social and vocational functioning (Victorian Department of Human Services, 2007). However, these government initiatives have largely overlooked offender’s unique needs, even though offenders have high levels of psychiatric morbidity and are more likely to experience social, personal and vocational dysfunction and require multi-modal treatment options.
Furthermore, since the Second National Mental Health Plan (1998‑2003) (Australian Health Ministers, 1998) there has been a move towards providing mental health services to people with high-prevalence disorders such as depression, anxiety and substance use. This initiative was introduced to overcome the public mental health services predominant focus on treating people with severe mental illnesses such as schizophrenia disorders. At the time, it was evident that the mental health system struggled to balance cost-effective management of patients with schizophrenia with the needs of other mentally ill individuals. Individuals with low-prevalence disorders, such as schizophrenia which affect approximately 1% of the general population, were utilising a disproportionately high amount of public mental health services due to having high treatment needs (Jablensky et al., 2000; Short et al., 2010). In contrast, people with high-prevalence disorders were unlikely to receive community mental health services as they are more likely to receive services from their general practitioner or private psychiatric services (Burgess et al., 2009).
Instead of investing in expanding community based mental health services to address this shortfall, the WHICH government introduced the Better Access program in 2006. The program aimed to address the low treatment rates among people with high-prevalence disorders by providing a rebate through the Medicare Benefits Schedule (MBS), for approved general practitioners, psychiatrists, psychologists, social workers and occupational therapists services (Department of Health and Ageing, 2008). Nonetheless, the Better Access program is not available in prisons. This has occurred even though prisoners with high-prevalence disorders are less likely to receive prison psychiatric services, as prison mental health services are geared to providing acutely unwell prisoners with low-prevalence disorders psychiatric treatment (Victorian Auditor-General’s Office, 2014). Therefore, prisoners with high-prevalence disorders are at a disadvantage once they enter prison as an opportunity to obtain psychiatric services which may substantially improve quality of life may be lost.
Australia has entered the third decade of targeted reforms of mental health services under the National Mental Health Strategy. Since the reforms commenced in the 1990s advocating major structural changes to the mental health system considerable changes have been achieved (Australian Health Ministers, 1992). Nonetheless, the changes have been inconsistent across jurisdictions and have mainly overlooked the unique and complex treatment needs of offenders.
Since deinstitutionalisation, public mental health services have not been able to keep up with demands of mentally ill individuals, especially mentally ill offenders (Victorian Auditor-General’s Office, 2014). The Australian mental health system currently has been criticised for operating at sub-optimal levels, as the system grapples with being able to afford comprehensive and effective services that matches the multifaceted needs and demands of the community (Meadows & Burgess, 2009; Short et al., 2010). Concerns have also been raised that a considerable number of people with mental illness do not receive mental health services (Burgess et al., 2009) and the system has failed to provide integrated care for people with a dual-diagnosis (Teesson et al., 2009). These are pertinent issues for prisoners, as the impact of mental health reforms on prisoners may be magnified. It is probable that the inadequacies of the mental health system impacts to a greater extent on the accessibility and utilisation of services by prisoners. This reflects, the higher rates of mental illness, including schizophrenia, as well as co-morbidity among prisoners, that in turn heightens the need for multi-modal psychiatric care.
It is likely that the underlying cause for the higher prevalence rates among prisoners is not attributed to one single factor, but instead results from a complex interaction of multiple factors. Irrespective of the cause, it is now widely accepted that prisoners constitute a vulnerable sub-group of the population that experiences mental illness at higher rates than the general population. Furthermore, since deinstitutionalisation, substantial inadequacies have developed in the provision of community mental services that may impact more on offenders than the general population. This reflects that offenders experience more psychosocial and environmental risk factors, as well as higher rates of substance abuse which may create barriers in obtaining services. National reforms and studies have also mainly excluded offenders even though there is a clear need for offenders to be incorporated. It would be arguably imprudent for mental health polices to be based solely upon data collected with the general population as this may further marginalise offenders who are an already highly-stigmatised group. A need was highlighted for empirical studies to investigate whether inadequacies of the mental health system impacts to a greater extent on accessibility of services for offenders.
Cite This Work
To export a reference to this article please select a referencing stye below:
Related ServicesView all
DMCA / Removal Request
If you are the original writer of this essay and no longer wish to have the essay published on the All Answers website then please: